Travel time generators are often used to calculate or predict the expected travel time between two locations. Current travel time generators typically use the actual street network represented in a graphic information system database to determine a route between two selected locations. Once the route is selected, the travel time generators determine the type of street and distance associated with each segment of the route. Additionally, the travel time generators assign a speed for the various street types associated with the route segments. The distance associated with each segment and the assigned speeds are then used to compute an expected travel time for each segment and for the overall route.
Although, current travel time generators provide a rough estimate of the expected travel time between two locations, they can be inaccurate. For example, the travel speed of a vehicle traveling in an urban area can vary significantly within each route segment based on the time of day, the direction of travel, and the geographic attributes of the area. Inaccuracy in travel time predictions can be problematic in many ways. For example, if the estimated travel time for a mass transit vehicle is too low, the vehicle can run behind schedule, resulting in driver and rider complaints. If the estimated travel time for a mass transit vehicle is too high, the result can be that the vehicle must slow down or wait at various locations, resulting in system wide inefficiencies.